little yellow-fringed flowers hung in
profusion from their spreading terminal racemes. I recalled their
singular shape, and the two outstretched stamens protruding from their
gaping corolla, and could distinctly see them as I sat in the carriage.
I had never chanced to read of this flower in the literature of
cross-fertilization, and murmuring, half aloud, "What pretty mystery is
yours, my Collinsonia?" prepared to investigate.

[Illustration: Fig. 9]

What I observed is pictured severally at Fig. 9, the flowers being
shown from above, showing the two spreading stamens and the decidedly
exceptional unsymmetrical position of the long style extending to the
side. A small nectar-seeking bumblebee had approached, and in alighting
upon the fringed platform grasped the filaments for support, and thus
clapped the pollen against his sides. Reasoning from analogy, it would
of course be absolutely clear that this pollen has thus been deposited
where it will come in contact with the stigma of another flower. So, of
course, it proved. In the bee's continual visits to the several flowers
he came at length to the younger blooms, where the forked stigmas were
turned directly to the front, while the immature stamens were still
curled up in the flower tubes. Even the unopened buds showed a number of
species where the early matured stigma actually protruded through a
tiny orifice in precisely the right position to strike the pollen-dusted
body of the bee, as he forced his tongue through the tiny aperture.[A]

[Footnote A: In numerous instances observed since the above was written
I have noted the larger bumblebees upon the blossom. These insects have
a different method of approach, hanging beneath the flower, the anthers
being clapped against their thorax at the juncture of the wings, instead
of the abdomen, as in the smaller bee.]

[Illustration]

If their dainty mechanism excite our wonder, what shall be said of the
revelations in the great order of the Compositae, where each so-called
flower, as in the dandelion, daisy, cone-flower, marigold, is really a
dense cluster of minute flowers, each as perfect in its construction as
in the examples already mentioned, each with its own peculiar plan
designed to insure the transfer of its own pollen to the stigma of its
neighbor, while excluding it from its own?

All summer long the cone-flower, Fig. 10 (_Rudbeckia hirta_), blooms in
our fields, but how few of us imagine the strange proce